Application of Pressurized Solvents for Ultrafast Trypsin Hydrolysisin Proteomics: Proteomics on the Fly

Daniel Lo´pez-Ferrer,

†

Konstantinos Petritis,

†

Kim K. Hixson,

‡

Tyler H. Heibeck,

†

Ronald J. Moore,

†

Mikhail E. Belov,

†

David G. Camp II,

†

and Richard D. Smith*

,†

Biological Sciences Division and Environmental Molecular Sciences Laboratory, Paciﬁc Northwest National Laboratory, P.O. Box 999, Richland, Washington 99352

Received December 3, 2007

A new method for rapid proteolytic digestion of proteins under high pressure that uses pressure cyclingtechnology in the range of 5

35 kpsi was demonstrated for proteomic analysis. Successful in-solutiondigestions of single proteins and complex protein mixtures were achieved in 60 s and then analyzedby reversed phase liquid chromatography-electrospray ionization ion trap-mass spectrometry. Methodperformance in terms of the number of

Shewanella oneidensis

peptides and proteins identiﬁed in ashotgun approach was evaluated relative to a traditional “overnight” sample preparation method.Advantages of the new method include greatly simpliﬁed sample processing, easy implementation,no cross contamination among samples, and cost effectiveness.

Keywords:

Mass spectrometry

•

trypsin digestion

•

pressure cycling technology

•

bottom-up proteomics

Introduction

“Omics” technologies, such as genomics, proteomics, me-tabolomics, and so forth, have not only provided informationneeded to understand relationships and interactions in biologi-cal systems,

1,2

but also contributed towards driving develop-ments that have increased analysis throughput. Key to further-ing these advances is a reduction in analysis time. An increasein the throughput of sample processing and analysis directly translates into increased experimental success in terms of depthof knowledge gained and decreased costs, which are importantto both academic and industrial realms.Mass spectrometry (MS)-based proteomics is a powerfultechnique for characterizing individual proteins or very complex protein mixtures, such as whole cell lysates. Theintegration of several key elements, such as improved MSinstrumentation,

3–5

multidimensional chromatographic sepa-rations,

computational biology, and signal processing appliedto MS data

has resulted in increasingly fast proteomicsanalyses and data processing; however, sample preparationtimes increasingly become a signiﬁcant bottleneck in theanalysis pipeline. While several sample preparation schemeshave been used in an attempt to increase throughput,

nonehave been adopted with universal appeal. Traditional strategiesinclude the enzymatic digestion of proteins either in-solutionor from gel spots after polyacrylamide gel electrophoresis(PAGE). In both strategies, the enzymatic digestion of theproteins is a critical and oftentimes a time-consuming step.Protein digestion has traditionally been performed using serine protease trypsin in a buffered medium over a deﬁnedlength of time, generally overnight (

∼

12 h), which presently makes protein digestion one of the most time-consuming stepsin the proteome analysis workflow. The success of a trypsindigestion is primarily defined by the time it takes to obtain acomplete and accurate proteolysis and the selectivity of theenzyme to access the reactive amino acid cleavage sites.Enzymatic reactions rates strongly depend on many environ-mental factors, including temperature, solvents used, pH range,and enzyme-to-substrate ratio. With regard to temperature,higher temperatures usually facilitate faster reaction rates; thatis, until the temperature is so high that it denatures the activesite of the enzyme. Attempts at creating a more thermo-stabletrypsin, in particular, are continuously under study in hopesof accelerating its digestion efficiency

in proteomics.For in-solution digestions, an increase in temperature may not be ideal since the heat can cause some proteins toaggregate and precipitate. Even if higher temperatures areavoided, a means of denaturing the proteins in order for trypsinto have access to the active cleavage sites on the proteins isstill necessary, and this has been traditionally achieved by theuse of chaotropes and/or surfactants. However, the use of thesechemical denaturants can present their own challenges, includ-ing the inactivation of enzyme activity or incompatibility withdownstream MS analysis. For example, when urea is used as achaotrope, the formation of ammonium cyanate can result inthe carbamylation of free amine groups on the proteins andpeptides.

Several works have been reported in which chemical dena-turants were replaced by solvent-assisted digestions.

The useof mixed-solvent buffers containing various concentrations of organic solvent (i.e., methanol, acetonitrile or isopropanol)resulted not only in protein denaturation, but also in better

* To whom correspondence should be addressed. R. D. Smith, PaciﬁcNorthwest National Laboratory, P.O. Box 999, MSIN: K8-98, Richland, WA 99352 (e-mial: rds@pnl.gov).

†

Biological Sciences Division, Paciﬁc Northwest National Laboratory.

‡

Environmental Molecular Sciences Laboratory, Paciﬁc Northwest Na-tional Laboratory.

10.1021/pr7008077 CCC: $40.75



XXXX American Chemical Society

Journal of Proteome Research

XXXX, xxx, 000

Published on Web 07/08/2008

solubilization of hydrophobic proteins.

Because these mixedaqueous

organic buffer systems can affect enzymatic proper-ties (due to the distortion of the protein solvation shell by theorganic solvent which can positively affect protein stability),optimizing these buffers has resulted in the achievement of higher efﬁciencies in the proteolytic process in terms of morepeptide identiﬁcations and shorter reaction times.

11,13,14

An-other important factor that dictates enzyme reaction rates isthe protease/substrate ratio. Higher enzyme concentrationsproduce faster reaction times, but at the expense of increasing the concentration of (unwanted) peptides produced by autoly-sis, which ultimately results in a loss of sensitivity.Combined optimization of the aforementioned parametershas proven to decrease digestion times.

Havlis et al. developeda procedure to reduce the trypsin digestion time to only 30min for proteins isolated by gel electrophoresis. The proceduremade use of a modiﬁed trypsin and was based on (1) a highertrypsin concentration in the digestion buffer, (2) an increasein the digestion temperature, which was performed at 57

C(due to the use of a thermo-stable trypsin) instead of the moretypical 30

C, and (3) the addition of organic solvents, whichhelped the enzyme gain access to the substrate cleavage sites.More recently, alternative energy inputs have been appliedto digestions to further increase enzyme reaction rates. Onealternative method involves the use of microwave energy, whichhastens enzymatic digestions to 3

5 min. This approach,referred to as microwave-assisted protein enzymatic digestion(MAPED) was useful for either in-solution

or in-gel

diges-tions and was shown to be applicable to complex proteinmixtures.

More recently, a new method

applies highintensity focused ultrasound (HIFU) to small sample volumes(typically 20

L) of protein in the presence of trypsin forultrafast protein digestion of either in-solution or in-gel proteinsamples. HIFU treatment was believed to act by boosting theenzyme

substrate kinetics due to the cavitation effects pro-duced during the ultrasound irradiation.

In-gel and in-solution protein digestions, using HIFU, were achieved in15

30 s.Considering the recent advances in sample preparation,particularly MAPED and HIFU, one optimization parameterthat has yet to be fully studied is the effect of pressure onenzymatic activity. High pressure processing (HPP) has beena well-known and established technology for many applicationsin the food and biotechnology industries. The majority of the work published in this area has been related to pressure foodprocessing to inactivate food pathogens and inhibit thoseenzymes that degrade food.

In addition, substantial evidenceexists to suggest that low to moderate pressures (e.g., 100

400MPa) can activate enzyme activity,

whereas higher pressuresinactivate the same enzymes.In the study reported herein, we have explored the applica-tion of pressure cycling technology (PCT) to prepare samplesfor proteomics analysis. PCT has been used in sample prepara-tion for many different analytical tasks and has been marketedas an efﬁcient method for cell lysis.

In our study, we evaluatedthe performance of PCT as a means of enhancing enzymeactivity. Using different organic solvent compositions, thesamples were subjected to PCT at different pressures and cyclecounts for 60 s. The pressurized trypsin digestion procedure was compared to a common overnight digestion to demon-strate proof of concept in a shotgun proteomics analysis.

Materials and Methods

Materials and Reagents.

Sequencing grade trypsin wasobtained from Promega (Madison,WI). Bovine serum albumin(BSA), myoglobin, urea, dithiothretiol (DTT), iodoacetamide(IAA), ammonium bicarbonate (Ambic), formic acid, and HPLCgrade solvents were purchased from Sigma-Aldrich (St. Louis,MO). The 0.1 mm zirconia/silica beads were ordered fromBiospec (Bartlesville, OK).

In-Solution Digestions.

Bovine serum albumin was used asa standard protein to evaluate the method under differentconditions. First, 6 mg of BSA was denatured in 8 M urea andreduced with 10 mM DTT in 25 mM ammonium bicarbonate(pH 8.25) at 37

C for 1 h. Iodoacetamide was added to a ﬁnalconcentration of 50 mM, and the resulting mixture wasincubated at room temperature in the dark for 45 min. Twelve50-

g aliquots were diluted 4-fold to reduce the urea concen-tration, using either 25 mM ammonium bicarbonate, 20%MeOH, or 80% MeOH. Trypsin was added (1:50 protease-to-protein ratio), to a ﬁnal volume of 1.4 mL and the solutions were placed in pulse tubes. The Barocycler NEP-3229 instru-ment and disposable polypropylene PULSE tubes FT-500 wereobtained from Pressure BioSciences (West Bridgewater, MA)and were used for all experiments. The pulse tubes weresubjected to the Barocycler program, using 4 or 8 pressurepulses for a total of 1 min per run. Finally, the enzymatic digests were transferred to new centrifuge tubes, acidiﬁed, and frozen with liquid N

to stop the reaction. The samples were then drieddown by centrifugal evaporation and stored at

C.The

Shewanella oneidensis

, strain MR-1, whole cell proteintryptic digest was prepared as described elsewhere

with minormodiﬁcations. Brieﬂy, cells were lysed by bead beating, using 0.1 mm zirconia/silica beads in a mini-bead beater (Biospec,Bartlesville, OK) for 180 s at 4500 rpm. The lysate was collectedand placed immediately on ice to inhibit proteolysis, thendenatured with 8 M urea, 25 mM ammonium bicarbonate, 10mM DTT, (pH 8), and incubated for 1 h at 37

C. Iodoacetamide was added to a ﬁnal concentration of 50 mM, and the resultantmixture was incubated for 45 min at room temperature in thedark. The mixture was diluted 4-fold and, following the additionof trypsin (1:50 protease-to-protein ratio), was incubated eitherovernight at 37

C or for 1 min using PCT at 35 kpsi. A solution with a ﬁnal concentration of 1

M protein in 12.5mM ammonium bicarbonate was prepared for the myoglobinexperiments. Trypsin was added and the samples were digested(1:50 protease-to-protein ratio) during the pressure cycles inthe Barocycler.

MS Analysis.

For BSA analyses, 500 fmol of the protein digest was analyzed by LC-MS/MS. An Agilent HPLC-Chip system wascoupled with a MSD Trap XCT Ultra ion trap (Agilent Tech-nologies, Santa Clara, CA). The Agilent auto sampler was usedto load the samples at 4

C. Separations were performed using a 40-nL enrichment column and 43 mm

m analyticalcolumn packed with 5

m ZORBAX 300SB C18 particles. A ﬂow rate of 1

L/min was employed for enrichment and 600 nL/min afterward. Peptides were eluted using a 5 min gradientfrom 5% to 90% Solvent B (0.5% formic acid, 90% acetonitrile;Solvent A, 0.5% formic acid in water/acetonitrile 97:3), with aseparation window of

∼

2 min. The total analysis time was 12min. Each sample was analyzed in triplicate. To prevent crosscontamination among different samples, a blank was runbetween each set of replicates.

research articles

Lo´pez-Ferrer et al.

B Journal of Proteome Research

•

Vol. xxx, No. xx, XXXX

The data were acquired in survey scans from 500 to 1600amu (3 microscans) followed by ﬁve data dependent MS/MSscans, using an isolation width of 3 amu, a normalized collisionenergy of 35%, and a dynamic exclusion period of 2 min. MS/MS data were analyzed using Spectrum Mill software againstan in-house FASTA database that contained

S. oneidensis

MR-1and BSA proteins. Spectra that matched to BSA were manually veriﬁed.For the complex protein mixture analysis, 2

g of the

S.oneidensis

digest were analyzed using a custom-built capillary LC system coupled online with a linear ion trap mass spec-trometer (LTQ; Thermo-Fisher, San Jose, A) with an in-housedeveloped ESI source.

23,24

The LTQ mass spectrometer wasoperated in a data-dependent MS/MS mode (

400

2,000),in which a full MS scan was followed by ten MS/MS scans,using a normalized collision energy of 35% with a dynamicexclusion of 1 min. Protein identiﬁcation was carried out using SEQUEST to deduce protein sequences from the

S. oneidensis

MR-1 genome sequence.

Database search parameters in-cluded a dynamic modification search (i.e., the presence andabsence of the modification was searched) for Met oxidationand a static search (i.e., presence of the modification wassearched only) for carbamidomethylation on Cys. Error ratesfor peptide identifications were calculated as reported previ-ously.

To study myoglobin folding, the protein was directly infusedby a syringe pump (KD Scientiﬁc, Holliston, MA) at 1

L/mineither with or without previous pressure treatment, into an Agilent TOF MS through an ESI interface developed in house.

MS data were recorded over an

range of 500

2500 at ascan rate of 1 scan/s.

Results and Discussion

Effect of Pressure Treatment on in-Solution EnzymaticHydrolysis.

PCT studies were performed using a Barocyclerinstrument, which utilizes changes in hydraulic pressure tomanipulate the samples under analysis within a closed system(i.e., the Pulse Tube). These tubes allow the volume of thesample to be altered during the pressure cycles by the move-ment of a small piston in each tube, which reacts to thepressure of hydraulic fluid (30% ethylene glycol) surrounding the Pulse Tube. Initial studies focused on investigating theeffect of pressure on enzymatic digestion. The behavior of en-zymes under pressure is complicated by a wide range of experimental conditions that can affect the reaction. For thisexperiment, we used BSA as a standard protein and trypsin asthe proteolytic enzyme. BSA is a globular protein with a largenumber of disulfide bridges that provide a high level of stability to its tertiary structure, making it very resistant to unfolding and thus to protein digestion. To combat this resistance tounfolding, the BSA samples were reduced and alkylated. Inaddition, all sample analyses were performed in a bufferedmedia of ammonium bicarbonate at pH 8.1, which is optimalfor trypsin activity in traditional (i.e., ambient pressure)protocols.Because several studies have shown at least a 3-fold increasein enzyme activity, which corresponds to an increase inpressure of at least 30-fold, we first analyzed the effect of enzyme activity in terms of protein proteolytic products at 5,10, 20, and 35 kpsi for 60 s at each pressure (Figure 1). Notethat this range was chosen based on the operating pressuresof the Barocycler. Additionally, pressures above 40 kpsi tendto destabilize protein structures with a consequential loss of activity.

The chromatograms in Figure 1a indicate that trypsin activity was not compromised at any of the pressures. However, onthe basis of the number of identiﬁed peptides (Figure 1b),digestion was not as complete at 5 kpsi as those achieved athigher pressures, even though the chromatograms at 5, 10, and20 kpsi are similar. Although chromatograms belonging to the35 kpsi samples look signiﬁcantly different as compared withthe others, manual inspection of the MS spectra showed thesame peptides between chromatograms. When pressure wasapplied to solutions that contained BSA in the absence of trypsin, no protein degradation products were observed, whichindicates that the pressure treatment itself did not cause proteinfragmentation (data not shown).

Effect of Pressure, Static Time, Number of Cycles andBuffer Composition Using Pressure Cycling Technology.

PCTcan disrupt tissues, cells, and cellular structures in buffers orother solutions;

as a result, it is frequently used for extracting proteins and nucleic acids in reaction tubes. Our goal was todetermine if multiple pressure cycles would inﬂuence trypsinactivity. To evaluate the inﬂuence of rapid cycling between highand low pressures, BSA was digested under pressure, using either 4 or 8 differential pressure cycles for a total of 60 s. Tofurther analyze the combined effect of pressure in the presenceof an organic solvent for a trypsin digestion, identical BSA protein aliquots were subjected to pressure-digestion at 35 kpsiin the presence of (1) ammonium bicarbonate, (2) an 80:20 (v/v) mixture of ammonium bicarbonate/methanol, and (3) a 20:80 (v/v) mixture of ammonium bicarbonate/methanol. Organicmedia have been widely used in proteomics applicationsinvolving enzymatic reactions,

11,14

especially those applied tomembrane proteins. The properties of enzymes in mixedorganic

aqueous solvent systems are influenced by factorssuch as protein structure, presence of phase interfaces, dielec-tric constants, and so forth, all of which contribute to theperformance of an enzyme in its biocatalytic system.The histogram in Figure 2a shows that nearly identical resultsin terms of the number of unique peptide identifications wereobtained for samples digested in comparable digestion buffers,regardless of the number of cycles. The chromatographicprofiles in Figure 2b are also very similar for comparablebuffers. A comparison of these results obtained at cyclic

Figure 1.

(a) LC-MS/MS chromatograms of high pressure assisteddigestions as a function of digestion pressure. High pressure in-solution digestions of BSA were performed at 5, 10, 20, and 35kpsi. (b) Histograms depict the number of unique identiﬁedpeptides for each method.

Rapid Proteolytic Hydrolysis Using Pressurized Solvents

research articles

Journal of Proteome Research

•

Vol. xxx, No. xx, XXXX